I've felt similarly about RSS for a while now--I've made a ton of attempts to build my giant collection of subscriptions but always just burn out on maintaining it. Another issue is when I try to get anyone even slightly non-technical to use RSS they bounce off immediately; it sadly just seems too complex/too much overhead for a large number of users.
I've been trying to build a site/app that adds some features mentioned in this post ("upvoting" based on views, tiktok-style video experience in the app, etc), but it's still very much a WIP and doesn't exactly fix the complexity problems yet. Still, I get encouraged seeing more projects like the OPs that hopefully bring about some sort of RSS resurgence.
> I've made a ton of attempts to build my giant collection of subscriptions but always just burn out on maintaining it.
Seems to me your problem lies in this part:
> giant collection
Don’t add so much that you can’t deal with it. Concentrate on infrequently updated sources. Any news website, for example, is too much and shouldn’t be in your reader. A small creator or YouTube channel from whom you want to see (almost) everything does go in.
If you ever feel overwhelmed, you have too many feeds and should remove every single one you don’t feel is absolutely valuable. Exceptions can be made if e.g. you were on vacation and never checked the reader. In that case, mark as read instead of removing.
If you ever find yourself regularly skipping the content from a feed without reading, remove it then and there. If you’re not consuming at least 80% (made up number, adapt to yourself) of posts, it does not belong in your feed reader.
Doubling down on this, get a Reader that let's you filter, and do so judiciously.
I have some feeds where I only allow thru items with specific categories. Others I have dozens of filters for "spam". And some I just had to give up on because they enshittified their feed to inhibit differentiation of junk/sponsored/spam content from real content.
Personally I just self-hosted a personal FreshRSS instance, but you can also get a lot of similar features from a paid Inoreader account.
> I've made a ton of attempts to build my giant collection of subscriptions but always just burn out on maintaining it.
RSS subscriptions aren't like Pokemon. You don't have to catch them all. One of the major selling points of RSS is that you can subscribe to sites that update infrequently so you get notified when they have a new update instead of checking the site manually and being disappointed that it hasn't updated in three weeks or whatever.
Adding a bunch of sites that update hundreds of times a day is a great way to DDOS your own attention span
I've has a similar experience with my own project that summarizes rss articles--the results have largely been pretty good, but I found using a "reasoning" model had much better results.
This is very similar to how I've approached classifying RSS articles by topic on my personal project[1]. However to generate the embedding vector for each topic, I take the average vector of the top N articles tagged with that topic when sorted by similarity to the topic vector itself. Since I only consider topics created in the last few months, it helps adjust topics to account for semantic changes over time. It also helps with flagging topics that are "too similar" and merging them when clusters sufficiently overlap.
There's certainly more tweaking that needs to be done but I've been pretty happy with the results so far.
I've been trying to build a site/app that adds some features mentioned in this post ("upvoting" based on views, tiktok-style video experience in the app, etc), but it's still very much a WIP and doesn't exactly fix the complexity problems yet. Still, I get encouraged seeing more projects like the OPs that hopefully bring about some sort of RSS resurgence.
[0] https://jesterengine.com